Merge pull request #1016 from Max-Health-Inc/develop

🧪 Auto-PR: Merge `develop` → `test`

Author: proxy-smart-releaser[bot]

.github/scripts/inferno-oauth-automation.js

@@ -1,3 +1,6 @@
+// SPDX-FileCopyrightText: Max Health Inc.
+// SPDX-License-Identifier: AGPL-3.0-or-later OR LicenseRef-Commercial
+
 /**
  * Inferno SMART App Launch STU2.2 Compliance Test Runner
  * 
@@ -535,8 +538,7 @@ async function runStandaloneLaunchTests(sessionId, browser) {
     return await waitForSimpleTestCompletion(sessionId, run.id, browser);
 
   } catch (error) {
-    console.error('Standalone Launch tests error:', error.message);
-    throw error;
+    return groupFailure('Standalone Launch', error);
   }
 }
 
@@ -571,10 +573,7 @@ async function runTokenIntrospectionTests(sessionId, browser) {
     return await waitForSimpleTestCompletion(sessionId, run.id, browser);
 
   } catch (error) {
-    console.error('Token Introspection tests error:', error.message);
-    // Don't throw — Token Introspection failure shouldn't block the pipeline
-    console.log('WARNING: Token Introspection tests failed but continuing...');
-    return null;
+    return groupFailure('Token Introspection', error);
   }
 }
 
@@ -669,9 +668,7 @@ async function runEhrLaunchTests(sessionId, browser) {
     return await waitForEhrLaunchCompletion(sessionId, run.id, browser);
 
   } catch (error) {
-    console.error('EHR Launch tests error:', error.message);
-    console.log('WARNING: EHR Launch tests failed but continuing...');
-    return null;
+    return groupFailure('EHR Launch', error);
   }
 }
 
@@ -856,10 +853,7 @@ async function runBackendServicesTests(sessionId) {
     return await waitForSimpleTestCompletion(sessionId, run.id);
 
   } catch (error) {
-    console.error('Backend Services tests error:', error.message);
-    // Don't throw — Backend Services failure shouldn't block the pipeline yet
-    console.log('WARNING: Backend Services tests failed but continuing...');
-    return null;
+    return groupFailure('Backend Services', error);
   }
 }
 
@@ -1009,6 +1003,17 @@ async function getSessionResults(sessionId) {
   return response.json();
 }
 
+/**
+ * Record a group that blew up, so the run can report every group's outcome and
+ * still fail. Three of the four groups used to log "WARNING: ... but continuing"
+ * and return null that nothing inspected, so a run reported success with most of
+ * the suite never executed; the fourth threw, which aborted the groups after it.
+ */
+function groupFailure(group, error) {
+  console.error(`${group} tests error:`, error.message);
+  return { group, ok: false, error: error.message };
+}
+
 async function printResults(results) {
   console.log('\n========================================');
   console.log('          TEST RESULTS SUMMARY          ');
@@ -1018,7 +1023,17 @@ async function printResults(results) {
   let failed = 0;
   let skipped = 0;
   let errors = 0;
-  
+  // 'omit' is Inferno declining a test that does not apply here (the TLS tests
+  // in local non-TLS mode). Not a failure, but it must be counted: it used to
+  // fall through to the default branch, inflating `total` while landing in no
+  // bucket at all.
+  let omitted = 0;
+  // A test that never reached a verdict: still 'wait'ing for a user action the
+  // automation never completed, or cancelled. These are the ones that made the
+  // job green while a launch flow was broken.
+  let incomplete = 0;
+  const incompleteTests = [];
+
   // Collect detailed failure info for later
   const failedTests = [];
 
@@ -1068,14 +1083,43 @@ async function printResults(results) {
           });
         }
         break;
+      case 'omit':
+        omitted++;
+        console.log(`− OMIT: ${title}`);
+        if (result.result_message) {
+          console.log(`    Reason: ${result.result_message.substring(0, 120)}`);
+        }
+        break;
+      case 'wait':
+      case 'cancel':
+        incomplete++;
+        incompleteTests.push({ title, status, test_id: result.test_id });
+        console.log(`⧗ INCOMPLETE (${status}): ${title}`);
+        break;
       default:
-        console.log(`? ${status.toUpperCase()}: ${title}`);
+        // An unrecognised status is counted as incomplete rather than ignored:
+        // a new Inferno state must not be able to pass the gate by being unknown.
+        incomplete++;
+        incompleteTests.push({ title, status, test_id: result.test_id });
+        console.log(`⧗ INCOMPLETE (unrecognised status "${status}"): ${title}`);
     }
   }
 
   console.log('\n========================================');
-  console.log(`Total: ${results.length} | Passed: ${passed} | Failed: ${failed} | Skipped: ${skipped} | Errors: ${errors}`);
+  console.log(`Total: ${results.length} | Passed: ${passed} | Failed: ${failed} | Incomplete: ${incomplete} | Skipped: ${skipped} | Omitted: ${omitted} | Errors: ${errors}`);
+  const accountedFor = passed + failed + skipped + errors + omitted + incomplete;
+  if (accountedFor !== results.length) {
+    console.log(`WARNING: ${results.length - accountedFor} result(s) landed in no bucket — the tally is wrong.`);
+  }
   console.log('========================================\n');
+
+  if (incompleteTests.length > 0) {
+    console.log('The following tests never reached a verdict:');
+    for (const t of incompleteTests) {
+      console.log(`  ⧗ [${t.status}] ${t.title}`);
+    }
+    console.log('');
+  }
 
   // Print detailed failure analysis
   if (failedTests.length > 0) {
@@ -1167,7 +1211,11 @@ async function printResults(results) {
     }
   }
 
-  return { passed, failed, skipped, errors, total: results.length };
+  return {
+    passed, failed, skipped, errors, omitted, incomplete,
+    total: results.length,
+    unaccounted: results.length - accountedFor,
+  };
 }
 
 async function main() {
@@ -1257,27 +1305,51 @@ async function main() {
     // Get all results across all groups
     const results = await getSessionResults(session.id);
     const summary = await printResults(results);
-    
+
+    const groupFailures = [standaloneResult, introspectionResult, backendServicesResult, ehrLaunchResult]
+      .filter(r => r && r.ok === false);
+
+    if (groupFailures.length > 0) {
+      console.error('\nTest groups that did not run to completion:');
+      for (const f of groupFailures) {
+        console.error(`  ✗ ${f.group}: ${f.error}`);
+      }
+    }
+
     // Output for GitHub Actions
     if (process.env.GITHUB_OUTPUT) {
       const fs = require('fs');
       fs.appendFileSync(process.env.GITHUB_OUTPUT, `passed=${summary.passed}\n`);
       fs.appendFileSync(process.env.GITHUB_OUTPUT, `failed=${summary.failed}\n`);
       fs.appendFileSync(process.env.GITHUB_OUTPUT, `total=${summary.total}\n`);
+      fs.appendFileSync(process.env.GITHUB_OUTPUT, `incomplete=${summary.incomplete}\n`);
+      fs.appendFileSync(process.env.GITHUB_OUTPUT, `omitted=${summary.omitted}\n`);
+      fs.appendFileSync(process.env.GITHUB_OUTPUT, `group_failures=${groupFailures.length}\n`);
     }
-    
-    // Exit with error if any tests failed OR no tests ran
-    if (summary.failed > 0 || summary.total === 0) {
-      console.error(`\n❌ Tests failed: ${summary.total === 0 ? 'No tests completed' : `${summary.failed} failed, ${summary.errors} errors out of ${summary.total}`}`);
+
+    // A run is only a pass if every test reached a verdict and every group ran.
+    // Counting only `failed` let a green result hide tests still stuck in `wait`
+    // because their launch flow never completed.
+    const reasons = [];
+    if (summary.total === 0) reasons.push('no tests ran');
+    if (summary.failed > 0) reasons.push(`${summary.failed} failed`);
+    if (summary.incomplete > 0) reasons.push(`${summary.incomplete} never reached a verdict`);
+    if (summary.unaccounted !== 0) reasons.push(`${summary.unaccounted} unaccounted for`);
+    if (groupFailures.length > 0) {
+      reasons.push(`${groupFailures.length} group(s) failed to run: ${groupFailures.map(f => f.group).join(', ')}`);
+    }
+
+    if (reasons.length > 0) {
+      console.error(`\n❌ Compliance run failed — ${reasons.join('; ')} (out of ${summary.total} tests, ${summary.errors} errors)`);
       process.exit(1);
     }
-    
+
     if (summary.errors > 0) {
       console.warn(`\n⚠️  ${summary.errors} test(s) had internal errors (not compliance failures) — ${summary.passed} passed out of ${summary.total}`);
     }
-    
-    console.log(`\n✅ All ${summary.passed} tests passed!`);
-    
+
+    console.log(`\n✅ All ${summary.passed} tests passed${summary.omitted > 0 ? ` (${summary.omitted} omitted as not applicable)` : ''}!`);
+
   } catch (error) {
     console.error('Test execution failed:', error.message);
     process.exit(1);

.github/workflows/smart-compliance-tests.yml

@@ -1317,17 +1317,20 @@ jobs:
           INFERNO_EXIT_CODE=${PIPESTATUS[0]}
           cd $GITHUB_WORKSPACE
           
-          # Parse results from the log
-          if grep -q "PASS:" inferno-output.log; then
-            PASSED=$(grep -c "✓ PASS:" inferno-output.log || true)
-            FAILED=$(grep -c "✗ FAIL:" inferno-output.log || true)
-          else
-            PASSED=0
-            FAILED=0
-          fi
-          PASSED=${PASSED:-0}
-          FAILED=${FAILED:-0}
-          
+          # Read the runner's own tally instead of re-counting log markers, which
+          # also matched the per-poll progress lines and could not see the tests
+          # that never reached a verdict.
+          SUMMARY_LINE=$(grep -E "^Total: [0-9]+ \| Passed:" inferno-output.log | tail -1 || true)
+          field() {
+            echo "$SUMMARY_LINE" | tr '|' '\n' | grep -iE "^ *$1:" | sed 's/.*: *//' | tr -d ' ' | head -1
+          }
+          TOTAL=$(field Total); TOTAL=${TOTAL:-0}
+          PASSED=$(field Passed); PASSED=${PASSED:-0}
+          FAILED=$(field Failed); FAILED=${FAILED:-0}
+          INCOMPLETE=$(field Incomplete); INCOMPLETE=${INCOMPLETE:-0}
+          OMITTED=$(field Omitted); OMITTED=${OMITTED:-0}
+          ERRORS=$(field Errors); ERRORS=${ERRORS:-0}
+
           # Write summary
           echo "" >> $GITHUB_STEP_SUMMARY
           echo "## SMART 2.2.0 Compliance Test Results" >> $GITHUB_STEP_SUMMARY
@@ -1341,16 +1344,31 @@ jobs:
           
           echo "" >> $GITHUB_STEP_SUMMARY
           echo "### Test Summary" >> $GITHUB_STEP_SUMMARY
+          echo "- Total: $TOTAL" >> $GITHUB_STEP_SUMMARY
           echo "- Passed: $PASSED" >> $GITHUB_STEP_SUMMARY
           echo "- Failed: $FAILED" >> $GITHUB_STEP_SUMMARY
+          echo "- Never reached a verdict: $INCOMPLETE" >> $GITHUB_STEP_SUMMARY
+          echo "- Omitted (not applicable here): $OMITTED" >> $GITHUB_STEP_SUMMARY
+          echo "- Internal errors: $ERRORS" >> $GITHUB_STEP_SUMMARY
           echo "- Test Stage: ${{ env.TEST_STAGE }}" >> $GITHUB_STEP_SUMMARY
           echo "" >> $GITHUB_STEP_SUMMARY
+
+          # Per-group marks were hardcoded to ✓, so the summary claimed all four
+          # groups passed even when three of them never ran. Derive them.
           echo "### Groups Tested" >> $GITHUB_STEP_SUMMARY
-          echo "- ✓ Standalone Launch (Discovery + OAuth + OIDC + Token Refresh)" >> $GITHUB_STEP_SUMMARY
-          echo "- ✓ Token Introspection" >> $GITHUB_STEP_SUMMARY
-          echo "- ✓ Backend Services (client_credentials + asymmetric JWT)" >> $GITHUB_STEP_SUMMARY
-          echo "- ✓ EHR Launch (EHR-initiated launch with patient context)" >> $GITHUB_STEP_SUMMARY
-          
+          group_mark() {
+            if grep -qF "✗ $1:" inferno-output.log; then echo "✗"; else echo "✓"; fi
+          }
+          echo "- $(group_mark 'Standalone Launch') Standalone Launch (Discovery + OAuth + OIDC + Token Refresh)" >> $GITHUB_STEP_SUMMARY
+          echo "- $(group_mark 'Token Introspection') Token Introspection" >> $GITHUB_STEP_SUMMARY
+          echo "- $(group_mark 'Backend Services') Backend Services (client_credentials + asymmetric JWT)" >> $GITHUB_STEP_SUMMARY
+          echo "- $(group_mark 'EHR Launch') EHR Launch (EHR-initiated launch with patient context)" >> $GITHUB_STEP_SUMMARY
+
+          if [ "$INCOMPLETE" -gt 0 ]; then
+            echo "" >> $GITHUB_STEP_SUMMARY
+            echo "> $INCOMPLETE test(s) never reached a verdict — see \"never reached a verdict\" in the log." >> $GITHUB_STEP_SUMMARY
+          fi
+
           # Include FHIR server info
           echo "" >> $GITHUB_STEP_SUMMARY
           echo "### FHIR Server Configuration" >> $GITHUB_STEP_SUMMARY

backend/package.json

@@ -1,7 +1,7 @@
 {
   "name": "proxy-smart-backend",
   "displayName": "Proxy Smart Backend",
-  "version": "0.3.13-beta.202608131856.7385093a0",
+  "version": "0.3.13-alpha.202608131916.083da3785",
   "type": "module",
   "scripts": {
     "test": "bun test --isolate",

config/eslint/package.json

@@ -1,6 +1,6 @@
 {
   "name": "@proxy-smart/eslint-config",
-  "version": "0.3.13-beta.202608131610.f7f7f64bd",
+  "version": "0.3.13-beta.202608131856.7385093a0",
   "private": true,
   "type": "module",
   "exports": {

deploy/infra/package.json

@@ -2,7 +2,7 @@
   "name": "proxy-smart-infra",
   "displayName": "Proxy Smart Infrastructure",
   "description": "AWS CDK infrastructure for Proxy Smart production deployment",
-  "version": "0.3.13-beta.202608131856.7385093a0",
+  "version": "0.3.13-alpha.202608131916.083da3785",
   "private": true,
   "type": "module",
   "scripts": {

frontend/smart-dicom-template/package.json

@@ -3,7 +3,7 @@
   "displayName": "SMART DICOM Algorithm Template",
   "description": "Starter kit for building SMART on FHIR imaging algorithm apps. Clone, implement your algorithm in src/algorithm.ts, and deploy as a SMART app.",
   "private": true,
-  "version": "0.3.13-beta.202608131610.f7f7f64bd",
+  "version": "0.3.13-beta.202608131856.7385093a0",
   "type": "module",
   "scripts": {
     "dev": "vite --port 5180",

frontend/ui/package.json

@@ -3,7 +3,7 @@
   "displayName": "Proxy Smart Admin UI",
   "description": "A web-based administration interface for managing healthcare applications and resources via Proxy Smart.",
   "private": true,
-  "version": "0.3.13-beta.202608131610.f7f7f64bd",
+  "version": "0.3.13-beta.202608131856.7385093a0",
   "type": "module",
   "scripts": {
     "dev": "vite",

package.json

@@ -1,6 +1,6 @@
 {
   "name": "proxy-smart",
-  "version": "0.3.13-beta.202608131856.7385093a0",
+  "version": "0.3.13-alpha.202608131916.083da3785",
   "repository": {
     "type": "git",
     "url": "git+https://github.com/Max-Health-Inc/proxy-smart.git"

packages/app-store/package.json

@@ -1,6 +1,6 @@
 {
   "name": "@proxy-smart/app-store",
-  "version": "0.3.13-beta.202608131610.f7f7f64bd",
+  "version": "0.3.13-beta.202608131856.7385093a0",
   "private": false,
   "type": "module",
   "description": "SMART on FHIR app store — manifest discovery, visibility configuration, and registry CRUD. Framework-agnostic.",

packages/auth/package.json

@@ -1,6 +1,6 @@
 {
   "name": "@proxy-smart/auth",
-  "version": "0.3.13-beta.202608131610.f7f7f64bd",
+  "version": "0.3.13-beta.202608131856.7385093a0",
   "private": false,
   "type": "module",
   "description": "SMART on FHIR STU 2.2.0 server-side authorization proxy — launch context, session management, token enrichment. Framework-agnostic, IdP-pluggable.",